[Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Andreas Färber]

This simplifies the loop and aids with refactoring of CPU list.

Requested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Andreas Färber <afaerber@suse.de>
---
 hw/cpu/a15mpcore.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/hw/cpu/a15mpcore.c b/hw/cpu/a15mpcore.c
index af182da..9abba67 100644
--- a/hw/cpu/a15mpcore.c
+++ b/hw/cpu/a15mpcore.c
@@ -50,7 +50,6 @@ static int a15mp_priv_init(SysBusDevice *dev)
     SysBusDevice *busdev;
     const char *gictype = "arm_gic";
     int i;
-    CPUState *cpu;
 
     if (kvm_irqchip_in_kernel()) {
         gictype = "kvm-arm-gic";
@@ -72,8 +71,8 @@ static int a15mp_priv_init(SysBusDevice *dev)
     /* Wire the outputs from each CPU's generic timer to the
      * appropriate GIC PPI inputs
      */
-    for (i = 0, cpu = first_cpu; i < s->num_cpu; i++, cpu = cpu->next_cpu) {
-        DeviceState *cpudev = DEVICE(cpu);
+    for (i = 0; i < s->num_cpu; i++) {
+        DeviceState *cpudev = DEVICE(qemu_get_cpu(i));
         int ppibase = s->num_irq - 32 + i * 32;
         /* physical timer; we wire it up to the non-secure timer's ID,
          * since a real A15 always has TrustZone but QEMU doesn't.
-- 
1.8.1.4

Re: [Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Andreas Färber]

Am 21.08.2013 18:45, schrieb Andreas Färber:
> This simplifies the loop and aids with refactoring of CPU list.
> 
> Requested-by: Peter Maydell <peter.maydell@linaro.org>
> Signed-off-by: Andreas Färber <afaerber@suse.de>
> ---
>  hw/cpu/a15mpcore.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)

Seeing no objection or pick-up from PMM I've applied this to qom-cpu
now, to move forward with CPU list refactoring.

https://github.com/afaerber/qemu-cpu/commits/qom-cpu

Andreas

> 
> diff --git a/hw/cpu/a15mpcore.c b/hw/cpu/a15mpcore.c
> index af182da..9abba67 100644
> --- a/hw/cpu/a15mpcore.c
> +++ b/hw/cpu/a15mpcore.c
> @@ -50,7 +50,6 @@ static int a15mp_priv_init(SysBusDevice *dev)
>      SysBusDevice *busdev;
>      const char *gictype = "arm_gic";
>      int i;
> -    CPUState *cpu;
>  
>      if (kvm_irqchip_in_kernel()) {
>          gictype = "kvm-arm-gic";
> @@ -72,8 +71,8 @@ static int a15mp_priv_init(SysBusDevice *dev)
>      /* Wire the outputs from each CPU's generic timer to the
>       * appropriate GIC PPI inputs
>       */
> -    for (i = 0, cpu = first_cpu; i < s->num_cpu; i++, cpu = cpu->next_cpu) {
> -        DeviceState *cpudev = DEVICE(cpu);
> +    for (i = 0; i < s->num_cpu; i++) {
> +        DeviceState *cpudev = DEVICE(qemu_get_cpu(i));
>          int ppibase = s->num_irq - 32 + i * 32;
>          /* physical timer; we wire it up to the non-secure timer's ID,
>           * since a real A15 always has TrustZone but QEMU doesn't.
> 


-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

Re: [Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Peter Maydell]

On 28 August 2013 16:19, Andreas Färber <afaerber@suse.de> wrote:
> Am 21.08.2013 18:45, schrieb Andreas Färber:
>> This simplifies the loop and aids with refactoring of CPU list.
>>
>> Requested-by: Peter Maydell <peter.maydell@linaro.org>
>> Signed-off-by: Andreas Färber <afaerber@suse.de>
>> ---
>>  hw/cpu/a15mpcore.c | 5 ++---
>>  1 file changed, 2 insertions(+), 3 deletions(-)
>
> Seeing no objection or pick-up from PMM I've applied this to qom-cpu
> now, to move forward with CPU list refactoring.

I wasn't sure how this patch fit with your larger series
that redoes the a15mpcore/gic/etc (which I think I'm
still waiting for an applyable patch set so I can review
it); I think I thought it was an apply-afterwards thing.

Anyway, I have no objection to it.

-- PMM

Re: [Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Andreas Färber]

Am 28.08.2013 17:22, schrieb Peter Maydell:
> On 28 August 2013 16:19, Andreas Färber <afaerber@suse.de> wrote:
>> Am 21.08.2013 18:45, schrieb Andreas Färber:
>>> This simplifies the loop and aids with refactoring of CPU list.
>>>
>>> Requested-by: Peter Maydell <peter.maydell@linaro.org>
>>> Signed-off-by: Andreas Färber <afaerber@suse.de>
>>> ---
>>>  hw/cpu/a15mpcore.c | 5 ++---
>>>  1 file changed, 2 insertions(+), 3 deletions(-)
>>
>> Seeing no objection or pick-up from PMM I've applied this to qom-cpu
>> now, to move forward with CPU list refactoring.
> 
> I wasn't sure how this patch fit with your larger series
> that redoes the a15mpcore/gic/etc (which I think I'm
> still waiting for an applyable patch set so I can review
> it); I think I thought it was an apply-afterwards thing.

This one needs to go before my qom-cpu-12 series to avoid the CPU loop
ugliness. That's in my CPU tree, and seeing no objections I still need
to test the p_greetings2.c sample for the CPU-removed case in linux-user
and will then merge that before we get more CPU loops.

For the MPCore series in my ARM/Tegra tree you wanted an assert for
object_initialize(). I am waiting for you to comment on my more
invasive/complete series (last part being optional), which would then
with any suggested changes go through my QOM tree before I add more uses
of object_initialize() that would need to be converted.

Based on that I will respin the MPCore, virtio and other QOM series.

> Anyway, I have no objection to it.

Thanks, want me to add some *-by?

Andreas

-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

Re: [Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Peter Maydell]

On 28 August 2013 16:35, Andreas Färber <afaerber@suse.de> wrote:
> For the MPCore series in my ARM/Tegra tree you wanted an assert for
> object_initialize(). I am waiting for you to comment on my more
> invasive/complete series (last part being optional), which would then
> with any suggested changes go through my QOM tree before I add more uses
> of object_initialize() that would need to be converted.

Ah, OK. I'm moving house this week so probably won't get to
detailed review til next week, but I think in general it looked OK.
Having to pass "&X, sizeof(X)" in the most common case is
a little unfortunate but I think the bus related code and so on
convinced me it was necessary.

-- PMM

Re: [Qemu-devel] [PATCH qom-cpu/arm-devs] hw/cpu/a15mpcore: Use qemu_get_cpu() for generic timers

[Andreas Färber]

Am 28.08.2013 18:11, schrieb Peter Maydell:
> On 28 August 2013 16:35, Andreas Färber <afaerber@suse.de> wrote:
>> For the MPCore series in my ARM/Tegra tree you wanted an assert for
>> object_initialize(). I am waiting for you to comment on my more
>> invasive/complete series (last part being optional), which would then
>> with any suggested changes go through my QOM tree before I add more uses
>> of object_initialize() that would need to be converted.
> 
> Ah, OK. I'm moving house this week so probably won't get to
> detailed review til next week, but I think in general it looked OK.
> Having to pass "&X, sizeof(X)" in the most common case is
> a little unfortunate but I think the bus related code and so on
> convinced me it was necessary.

Yeah, it turned into more work than I expected ;) but there were at
least two cases (still need to reply) where the macro-fied version
would've checked the wrong thing and my patches are intentionally
deviating from the usual pattern - main_system_bus and a sec_bus.

Thanks,
Andreas

-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg